Media Effects on Attitudes towards Immigration. A Multilevel, Multimethod, Longitudinal Comparative Study

经过几十年的内部和外部移民，欧洲社会在种族和文化上比当代历史上任何时候都更加多样化。这一趋势在公众和政治上都存在争议，有可能加剧政治两极分化和激进化，从而威胁到自由民主国家。许多人对移民缺乏可观的个人经验。对大多数公民来说，间接经验（主要由媒体传递）是对该问题形成态度的最相关来源。因此，媒体在塑造舆论方面起着重要的作用。最近关于这一问题的研究证明，媒体对移民的信仰和态度存在影响。然而，在理论和实证方面仍有很大的改进空间。在群体间威胁理论和属性议程设置方法的基础上，我们将对媒体在塑造欧洲社会对移民的信念和态度方面的影响进行方法学上的合理调查。我们通过探索媒体报道的变化如何与国民对移民的不同信念有关，以及这些信念如何影响对不同形式移民的态度来做到这一点。为了实现这一目标，我们采用多方法（3种方法）和多层次（2个水平，个人和国家）的设计，从纵向（18年）和国际比较的角度（6个国家）分析了媒体的影响。我们将使用两个公开可用的数据源：a)来自欧洲社会调查（ESS）的调查数据；b)来自欧盟统计局、Genesis和Eurobarometer的外生汇总数据。此外，我们将c)通过对报纸文章的自动内容分析来收集媒体内容数据。因此，我们将通过神经机器翻译（NMT）工具将德语、西班牙语和法语发表的文章翻译成英语，并在所有国家的所有报纸上应用英语代码本。将使用监督机器学习（SML）技术来执行属性分析。结果将增加我们对媒体报道在有争议的社会领域的影响的理解。此外，该研究将对宏观层面的媒体在不同时间和国家条件下对个人的影响进行实证检验，假设这些条件会影响媒体报道和个人的信仰和态度。从理论角度来看，这些发现将有助于完善所提出的情境化媒体效应模型。

After decades of internal and external migration, European societies are more ethnically and culturally diverse than ever before in contemporary history. This trend is publicly and politically disputed and has the potential to intensify political polarization and radicalization and thereby threaten liberal democracies. Many people lack appreciable personal experience with migration. For most citizens indirect experiences – mainly transferred by the media – are the most relevant source for attitude formation on the issue. Therefore, the media plays an important role in shaping the public opinion. Recent studies on the subject prove the existence of media influences on beliefs and attitudes towards migration. However, there is still plenty of room for theoretical and empirical improvements. Building on intergroup threat theory and the attribute agenda setting approach, we will provide a methodologically sound investigation on the influence of media in shaping beliefs about and attitudes on immigration in European societies. We do that by exploring how variations in media coverage relate to different dimensions of beliefs about immigration by nationals and how these beliefs influence the attitudes towards different forms of immigration. To achieve this goal, we analyze media effects in a longitudinal (18 years) and international comparative perspective (6 countries), applying a multimethod (3 methods) and multilevel (2 levels, individual and country) design. We will use two publicly available data sources: a) survey data from the European Social Survey (ESS)) and b) exogenous aggregate data from Eurostat, Genesis, and Eurobarometer. Furthermore, we will c) gather media content data by automated content analyses of newspaper articles. Therefore, we will translate articles published in German, Spanish, and French to English by neural machine-translation (NMT) tool, and apply an English codebook on all newspapers from all countries. Supervised machine learning (SML) technique to perform attribute analysis will be used. The results will add to our understanding of the influence of media coverage in a contested social sphere. Moreover, the research will provide an empirical test of macro-level media effects on individuals in the context of varying temporal and national conditions that are hypothesized to influence both the media coverage and individual beliefs and attitudes. From a theoretical perspective, the findings will help to refine the proposed model of contextualized media effects.